Review of random effects estimation with application to the growth of Acacia senegal trees

Abstract

Estimation of random effects (variance components) IS a method often used in population genetics and applied to areas such as animal and plant bleeding and growth. Scientists nowadays feel lost if confronted with huge set of different random effects estimation methods. This is especially because there exists no uniformly best method hence deciding which method should be used is difficult to take. This paper gives an overview of maximum likelihood and restricted maximum likelihood methods of estimating random effects applied to random coefficient models and demostrates them by applying them in determining variability in the growth of Acacia Senegal trees. We can say that both methods gives similar results with the dataset used. However random effects estimated using maximum likelihood method are slightly less than those estimated using restricted maximum likelihood methods. Random intercept and slope model is more appropriate to use in determination of variability in growth of Acacia Senegal trees than random intercept model. Keywords: Maximum Likelihood, Restricted Maximum Likelihood, Random Effects, Random Intercept Model, Random Intercept and Slope Model.